Detection And Identification Of Mortar Void In Ballastless Track Based On Elastic Wave Method

CRTS ? ballastless track carries the operation of high-speed trains,which is a hardware structure to ensure the safe running of high-density and high-speed trains.It can be divided into the composed of track plate,cement emulsified asphalt mortar layer(CA layer),supporting layer and so on.Under the effection of natural repeated environment load and train load,the CA mortar layer,as the elastic adjustment layer of ballastless track structure,gradually fails,deteriorates and forms void.If it is not detected and identified in time,it will further evolve into a disease,and even threaten the integrity and continuity of ballastless track structure and the safety of high-speed railway train operation.However,the void of CA mortar layer is located in the interior of ballastless track structure,and it can not be detected in time through its appearance.Aiming at detecting the void CA mortar layer in railway engineering,in this paper focuses on the key problems in elastic wave detection and identification methods,such as elastic wave propagation characteristics,disease echo characteristic parameter extraction,elastic wave signal classification and recognition,etc.(1)Propagation characteristics of elastic wave in ballastless track structureBased on the elastic wave propagation theory,the variation law of elastic wave conversion wave energy and reflection coefficient at the interface between ballastless track plate with the mortar layer is studied.When the elastic wave is incident on the interface of the normal ballastless track structure at a certain angle,the energy of the transmitted P wave at the interface is higher,and when the P wave is incident,the energy accounts for 80%-60% of the energy.When the S wave is incident,the highest proportion of energy is up to 50%.When the mortar layer is deteriorated,the energy of reflected P wave increases gradually.When the elastic wave is vertically incident,there are only reflected P wave and transmitted P wave in the structure,and the incident wave frequency affects the reflection coefficient of P wave.When the mortar layer is deteriorated,the reflection coefficient of high frequency incident wave can reach 98%.Based on the finite element numerical model of ballastless track with void of mortar layer created in this paper.the propagation characteristics of low frequency elastic wave,R wave and P wave in the layered structure of ballastless track are studied.If the size of cavitation damage in CA mortar layer is different,the reflection duration of all kinds of elastic wave is different.the small size void has less interference to the propagation path of elastic wave,and the reflection time is short.When the void size increases,the hindrance to the propagation of elastic wave increases and prolongs the propagation period of elastic wave in the structure.(2)Feature extraction and cassification of disease echo based on response characteristics of low frequency elastic waveBased on the results of the low frequency response in the finite element numerical model,the disease characteristic parameters of the low frequency response are constructed,and the effective characteristic parameters are analyzed and extracted through the model test.When there is no void in the mortar layer,the admittance curve is smooth and continuous,the difference between the admittance peak and valley is small,and there is no obvious peak at high frequency.If the void of the mortar layer is more than 0.3m,the greater the difference between the admittance peak and valley is,the more severe the oscillation of the admittance curve is.The One-class SVM algorithm is used to classify and identify the low frequency elastic wave signals.Through the optimization of the characteristic parameters of the disease echo and the comparison and selection of reasonable classification models,the nonlinear mapping relationship between the characteristic parameters with the void above 0.3m is established,and the high-precision classification of low-frequency elastic wave signals is investigated.(3)Feature extraction of disease echo based on high frequency surface waveBased on the surface wave response of the finite element numerical model and the dispersion results calculated based on the layered medium matrix transfer algorithm,the disease echo characteristic parameters of the surface wave response are optimized,and the effective characteristic parameters are analyzed and extracted through the model test.When the void size increases from 0.05 m to 0.2m,there will be a large jump in the high frequency components of the Fmurk spectrum,the dispersion energy group and the dispersion curve.The 0.05 m void absorbs the part of the energy and reduces the energy amplitude of the receiver.The reflection effect of the void with size of 0.1m and 0.2cm is stronger,which increases the energy amplitude.(4)Feature extraction and recognition of disease echo based on P waveBased on the P-wave response results of the finite element numerical model,the disease characteristic parameters of P-wave response are constructed,and the effective characteristic parameters are analyzed and extracted through the model test.By viewing the various frequencies of the waveforms recorded under different amplitudes in the frequency domain,the void degree of the mortar layer can be identified according to the frequency peak at different frequency bands,the distribution characteristics of formants in frequency domain and the value of power density.Wavelet analysis and Burg power spectrum analysis are used to identify and locate the hollow disease echo of P wave signal more accurately.Both of the two methods can enhance the weak echo signal in a specific frequency band of elastic wave,and the Burg power spectrum is more suitable for multiple signals with small frequency difference.In the Burg power spectrum of P wave signal,the echo of mortar layer and the echo of defect edge can be clearly distinguished,and the position of the boundary point of 0.05 m void can be accurately identified.